install.md

Standalone compilation of SOLVERLAB from sources

Prerequisites for the compilation of SOLVERLAB

The following package list is sufficient on Ubuntu 20.04 :

• cmake3 (mandatory)
• g++ or another C++ compiler (mandatory)
• python3-dev, python3-numpy and swig3for python scripts (mandatory)
• pyqt5-dev-tools to generate the Graphical User Interface (optional)
• python3-matplotlib, paraview-dev, libnetcdf-dev (on Ubuntu 20.04) and python3-paraview for postprocessing tools such as plotting curves (matplotlib) or generating 3D view images (paraview) (optional)
• ffmpeg and ffmpeg-devel to generate an animation from a set of curves (optional)
• libmpich-dev or libopenmpi-dev for mpi parallelisation (optional)
• libcppunit-dev, if you want to generate unit tests. Use the compilation option -DSOLVERLAB_WITH_TESTS=ON (optional).
• rpm, if you want to generate RPM installation packages. Use the compilation option -DSOLVERLAB_WITH_PACKAGE=ON (optional).

Download SOLVERLAB sources

Download SOLVERLAB sources from GitHub

• use the following url in a browser : https://github.com/ndjinga/SOLVERLAB/archive/master.zip, then download and unzip the file in a directory SOLVERLAB-master
• or type the following in a terminal : wget https://github.com/ndjinga/SOLVERLAB/archive/master.zip, then unzip the file in a directory SOLVERLAB-master
• or clone the git repository to a folder SOLVERLAB-master: git clone https://github.com/ndjinga/SOLVERLAB.git SOLVERLAB-master

Compile and install SOLVERLAB from source files

First create a directory named 'build' where the compilation will take place and open a terminal in that directory.

Simple configuration for a minimum version:
Type the following cmake instruction in a terminal for a minimal version

• cmake /path/to/SOLVERLAB-master/ -DCMAKE_INSTALL_PREFIX=../SOLVERLAB_install -DCMAKE_BUILD_TYPE=Release -DSOLVERLAB_WITH_DOCUMENTATION=ON

This will download and build the following dependencies

• PETSc from https://ftp.mcs.anl.gov/pub/petsc/release-snapshots/petsc-3.19.6.tar.gz
• SLEPc from https://slepc.upv.es/download/distrib/slepc-3.19.2.tar.gz
• F2CBLASLAPACK from https://ftp.mcs.anl.gov/pub/petsc/externalpackages/f2cblaslapack-3.8.0.q2.tar.gz
• HDF5 from https://support.hdfgroup.org/ftp/HDF5/releases/hdf5-1.10/hdf5-1.10.3/src/hdf5-1.10.3.tar.gz
• MEDFILE from http://git.salome-platform.org/gitpub/modules/med.git
• MEDCOUPLING from http://git.salome-platform.org/gitpub/tools/medcoupling.git
• CONFIGURATION from http://git.salome-platform.org/gitpub/tools/configuration.git

Advanced configuration for a complete version:
If you already have an installation of PETSC, MED and MEDCoupling, you may save computational time and memory by typing the following cmake instruction in a terminal :

• cmake /path/to/SOLVERLAB-master -DCMAKE_INSTALL_PREFIX=../SOLVERLAB_install -DCMAKE_BUILD_TYPE=Release -G"Eclipse CDT4 - Unix Makefiles" -D_ECLIPSE_VERSION=4.3 -DSOLVERLAB_WITH_DOCUMENTATION=ON -DPETSC_DIR=${PETSC_DIR} -DPETSC_ARCH=${PETSC_ARCH} -DMEDFILE_ROOT_DIR=${MEDFILE_ROOT_DIR} -DMEDCOUPLING_ROOT_DIR=${MEDCOUPLING_ROOT_DIR}

This assumes that you have an existing

• installation of PETSc (with submodule SLEPC) at the location given by the environment variable PETSC_DIR and the architecture variable PETSC_ARCH
  See the instructions given in the official documentation
• installation of MED at the location given by the environment variable MEDFILE_ROOT_DIR
• installation of MEDCOUPLING at the location given by the environment variable MEDCOUPLING_ROOT_DIR

The 2 dependencies MED and MEDCOUPLING should have been compiled with the same version of HDF5
Warning : the linux package libhdf5-dev is generally not compatible with the libraries MED and MEDCoupling

Compile and install:

• make
• make install

Run tests and documentation

Run unit and example tests:

• make examples

Run validation tests:

• make validation

Generate html user guide

• make doc-user

Generate html developer guide

• make doc

Parallel version

To compile a parallel version of SOLVERLAB, you first need parallel C and C++ compilers and a parallel installation of the packages PETSc, MED and MEDCoupling.

Then you need to add the following extra CMAKE parameters to the configuration step :

• -DMPI_HOME=${MPI_HOME} -DCMAKE_C_COMPILER=${MPI_C_COMPILER} -DCMAKE_CXX_COMPILER=${MPI_CXX_COMPILER}

where

• MPI_HOME stores the location of the mpi library
• MPI_C_COMPILER stores the location of the parallel C compiler
• MPI_CXX_COMPILER stores the location of the parallel C++ compiler

*/